Why I Ended Up in the Harness

2024-05-13 · Source: The Business Engineer · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Robotics & Autonomous Systems · Depth: Intermediate, long

Summary

Gennaro Cuofano details a forced professional transformation driven by the rapid evolution of AI from 2020 to 2026, characterized by four distinct scaling eras. Initially, the "pre-training scaling" era (2020-2024) focused on larger models and data, leading to chat interfaces and rewarding operating skill. By September 2024, the frontier shifted to "test-time reasoning," where models like OpenAI's o1 and Anthropic's reasoning Claude spent more compute on inference, expanding the unit of work from a turn to a task. The "agent era" in 2025 emerged with the Model Context Protocol (MCP), enabling autonomous agents to execute multi-step goals, with MCP seeing 97 million SDK downloads by late 2025. Currently, the "orchestration/swarms" era (2026 onwards) involves composing specialist agents into parallel swarms, shifting the business model to AGaaS and the form factor to ambient surfaces. This accelerated evolution commoditizes prior skills, compelling professionals to move from performing tasks to authoring outcomes.

Key takeaway

For AI/ML Directors evaluating team productivity and future strategy, recognize that AI's rapid scaling has commoditized traditional operating skills. You should transition your teams from manual execution to framing and directing agentic systems, leveraging AGaaS for outcome-based value. Failing to adapt means your team's output will fall below market expectations, as the baseline for individual productivity has fundamentally shifted. Embrace authorship and responsibility, not just capability.

Key insights

AI's rapid scaling shifts from models to orchestration, commoditizing prior skills and forcing a move from operating to authoring.

Principles

• AI scaling axes continually relocate outward, commoditizing prior skills.
• Each scaling shift creates new work units, business models, and form factors.
• Authorship (wanting, choosing, liability) remains non-commoditizable.

In practice

• Shift focus from operating AI to framing and directing agentic systems.
• Embrace AGaaS models for outcome-based value delivery.
• Prepare for AI-native interfaces beyond traditional screens.

Topics

• AI Scaling Laws
• AI Agents
• Agent Orchestration
• Business Model Transformation
• Future of Work
• Model Context Protocol

Best for: AI Architect, Entrepreneur, Director of AI/ML, AI Product Manager, Consultant

Related on AIssential

• Judgment and creativity are all you need. — Coding agents move software development bottlenecks from implementation to judgment and creativity.
• How to Reorg After AI Changes Everything | Block's Owen Jennings on the a16z Show — Advanced AI models fundamentally alter software development and organizational structures, enabling massive productivity gains.
• Fragments: June 2 — AI's impact spans metrics, job markets, model capabilities, content integrity, security, code quality, and human workflow, presenting both opportunities and complex challenges.
• Steve Yegge on AI Agents and the Future of Software Engineering — AI is rapidly transforming software development, making traditional coding skills less critical while increasing engineer productivity.
• What Happens When a Public Company Goes All In on AI — AI's rapid advancement fundamentally broke the "headcount equals output" law, enabling drastic productivity gains and organizational restructuring.
• 🔮 Exponential View #561: Token scaling; frontier revenues; Spotify engineers; century bonds, T-cells vs Alzheimer’s & glass replaces silicon++ — Rapid AI adoption is transforming individual workflows and enterprise operations, necessitating a re-evaluation of AI's role.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Business Engineer.